LogNormal Particles Size Distribution for Spray Simulations

cfd seeker · November 7, 2016, 17:16

Hello,

I am doing the Spray Simulations of GDI Engine in CFX and for that i need to provide some estimate of the particles size distribution. The particle size data which is available to me is Lognormal PDF(Probability density function) data for particle sizes as shown in the attached image. But in CFX the options which I could find are "Normal in Diameter by Number or Normal in Diameter by Mass" so these options don't fit my data as they will generate Normal distribution. I am wondering if there is a way in CFX to specify the lognormal PDF data in the text file? or does anybody has any other solution to this problem.

Any help would be appreciated. Thanks in advance.

Regards

ghorrocks · November 7, 2016, 17:36

The options I see in V17.2 are:

* Specified Diameter
* Uniform in Diameter by Number
* Uniform in Diameter by Mass
* Normal in Diameter by Number
* Normal in Diameter by Mass
* Rosin Rammler
* Nukiyama Tanasawa
* Discrete Diameter Distribution

Hopefully you should be able to fit one of those options to your required distribution.

Also note that the "Normal in diameter by mass" option will not give a normal distribution when viewed by number. You would have to calculate what distribution this gave when viewed by number, but it may be close to your required distribution as well.

cfd seeker · November 7, 2016, 18:01

Quote:

Originally Posted by ghorrocks

The options I see in V17.2 are:

* Specified Diameter
* Uniform in Diameter by Number
* Uniform in Diameter by Mass
* Normal in Diameter by Number
* Normal in Diameter by Mass
* Rosin Rammler
* Nukiyama Tanasawa
* Discrete Diameter Distribution

Hopefully you should be able to fit one of those options to your required distribution.

Thanks Gorrocks for your reply. I have seen and carefully studied all these options and I mentioned only two as those were little similar to my case.

Quote:

Also note that the "Normal in diameter by mass" option will not give a normal distribution when viewed by number. You would have to calculate what distribution this gave when viewed by number, but it may be close to your required distribution as well.

Yes you are right it will not give Normal distribution as at large sizes only smaller no. of particles will be required and at smaller sizes larger number of particles will be required. The quick guess which I can make it will give some kind of exponential function with negative slope but I don't think so it will fit my data. Further suggestions will be highly appreciated.

Do you or any other any body has idea about the parameters of Rosin Rammler and Nukiyama Tanasawa distributions. The Injector has six holes with a hole diameter of 0.193mm and fuel(iso-octane) is injected into the domain at 10MPa pressure at mass flow rate of 0.0136667k kg/s. Any help would be appreciated.

Regards

ghorrocks · November 7, 2016, 18:07

CFX does not have a lognormal distribution built in so you won't find a "click the button" option to do what you want. You are going to have to look at the distribution shapes of the available options and adjust them to get something similar to your intended distribution.

So I would look at all the options and see what distributions they give. The lognormal distribution is a reasonably simple distribution so you should be able to get something to be close enough that the difference is not important.

All I can suggest is that the "Normal in diameter by mass" option, also Rosin Rammler, Nukiyama Tanasawa and Discrete Diameter Distributions are the ones I would look at more closely. They all have several tunable constants so you are going to have to play with them to see what fits the best.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[ICEM] Help with fixing imported IGES model	siw	ANSYS Meshing & Geometry	24	August 24, 2010 12:22
Phase locked average in run time	panara	OpenFOAM	2	February 20, 2008 15:37
Plot particles w. correct size?	Michael B	CFX	1	August 8, 2006 03:36
particle size distribution	Felix	CFX	0	May 29, 2006 05:16
how to make particle size distribution	adma	CFX	0	January 10, 2006 16:57

November 7, 2016, 17:36		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,872 Rep Power: 144	The options I see in V17.2 are: * Specified Diameter * Uniform in Diameter by Number * Uniform in Diameter by Mass * Normal in Diameter by Number * Normal in Diameter by Mass * Rosin Rammler * Nukiyama Tanasawa * Discrete Diameter Distribution Hopefully you should be able to fit one of those options to your required distribution. Also note that the "Normal in diameter by mass" option will not give a normal distribution when viewed by number. You would have to calculate what distribution this gave when viewed by number, but it may be close to your required distribution as well.

November 7, 2016, 18:07		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,872 Rep Power: 144	CFX does not have a lognormal distribution built in so you won't find a "click the button" option to do what you want. You are going to have to look at the distribution shapes of the available options and adjust them to get something similar to your intended distribution. So I would look at all the options and see what distributions they give. The lognormal distribution is a reasonably simple distribution so you should be able to get something to be close enough that the difference is not important. All I can suggest is that the "Normal in diameter by mass" option, also Rosin Rammler, Nukiyama Tanasawa and Discrete Diameter Distributions are the ones I would look at more closely. They all have several tunable constants so you are going to have to play with them to see what fits the best.